--------------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  /Users/weiss.mehrabi/Desktop/Replication/replication_log.log
  log type:  text
 opened on:   9 Nov 2025, 21:23:28

. 
. 
. ********************************************************************************
. * 1. LOAD REPLICATION DATA
. ********************************************************************************
. 
. * Load trimmed replication dataset
. use "fpa_dataset.dta", clear

. 
. 
. ********************************************************************************
. * 2. TABLE 1: SUMMARY STATISTICS
. ********************************************************************************
. 
. * Generate and store summary statistics
. estpost summarize early_withdrawal fcoup scoup lbestdead coal1_state coal2_un ///
>     coal3_alliance coalsize direct2 domdispute combat ///
>     ltdistance lldistance logcap troops ///
>     newll election dem dura time time2 time3 ///
>     if type_of_conflict != 1 & 2 & polity2 < 7

             |  e(count)   e(sum_w)    e(mean)     e(Var)      e(sd)     e(min)     e(max)     e(sum) 
-------------+----------------------------------------------------------------------------------------
early_with~l |       203        203   .0591133   .0558943   .2364197          0          1         12 
       fcoup |       203        203   .0394089   .0380432   .1950467          0          1          8 
       scoup |       203        203   .0147783    .014632   .1209628          0          1          3 
   lbestdead |       203        203   1.014991   7.350884   2.711251          0   13.11735   206.0431 
 coal1_state |       203        203   .6403941   .2314295   .4810712          0          1        130 
    coal2_un |       203        203   .2118227   .1677803     .40961          0          1         43 
coal3_alli~e |       203        203   .1477833   .1265668   .3557623          0          1         30 
    coalsize |       203        203   5.679803   17.63459   4.199356          1         12       1153 
     direct2 |       203        203   .1871921   .1529045     .39103          0          1         38 
  domdispute |       203        203   .3546798   .2300151   .4795989          0          1         72 
      combat |       203        203   .7192118   .2029459   .4504952          0          1        146 
  ltdistance |       203        203   5.573285   10.95583    3.30996          0   9.130648   1131.377 
  lldistance |       203        203   6.913568   7.721106   2.778688          0   9.130648   1403.454 
      logcap |       203        203  -6.332526   1.309686   1.144415  -9.152071  -2.757172  -1285.503 
      troops |       203        203   2.098522    .297176   .5451385          1          3        426 
       newll |       203        203   .0689655   .0645271   .2540219          0          1         14 
    election |       203        203   .2413793   .1840219   .4289777          0          1         49 
         dem |       203        203    .044335   .0425791   .2063471          0          1          9 
        dura |       203        203   .5910461   1.543185    1.24225  -3.502769   2.239943   119.9824 
        time |       203        203   2.487685   3.508511   1.873102          1         11        505 
       time2 |       203        203   9.679803   272.7138   16.51405          1        121       1965 
       time3 |       203        203   53.14778   22253.43   149.1758          1       1331      10789 

. 
. * Export summary statistics to LaTeX
. esttab using "summary_stats_pervar.tex", ///
>     cells("count(fmt(0)) mean(fmt(2)) sd(fmt(2)) min(fmt(2)) max(fmt(2))") ///
>     label booktabs nonumber nomtitle unstack ///
>     title("Summary Statistics") ///
>     replace
(file summary_stats_pervar.tex not found)
(output written to summary_stats_pervar.tex)

. 
. 
. ********************************************************************************
. * 3. TABLE 2: MAIN MODELS (MODELS 1–7)
. ********************************************************************************
. 
. * Clear any stored estimates
. eststo clear

. 
. * Model 1: Liberal democracies not excluded
. logit early_withdrawal fcoup scoup lbestdead ///
>     if type_of_conflict != 1 & 2, cluster(intid) nolog

Logistic regression                                     Number of obs =    395
                                                        Wald chi2(3)  =  38.75
                                                        Prob > chi2   = 0.0000
Log pseudolikelihood = -50.049556                       Pseudo R2     = 0.2861

                                     (Std. err. adjusted for 43 clusters in intid)
----------------------------------------------------------------------------------
                 |               Robust
early_withdrawal | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
-----------------+----------------------------------------------------------------
           fcoup |   4.954196   1.147577     4.32   0.000     2.704986    7.203405
           scoup |   3.270223   1.236916     2.64   0.008     .8459117    5.694534
       lbestdead |    .202433   .1037324     1.95   0.051    -.0008787    .4057448
           _cons |   -3.96337    .446325    -8.88   0.000    -4.838151   -3.088589
----------------------------------------------------------------------------------

. eststo m1

. 
. * Model 2: Liberal democracies excluded
. logit early_withdrawal fcoup scoup lbestdead ///
>     if type_of_conflict != 1 & 2 & polity2 < 7, cluster(intid) nolog

Logistic regression                                     Number of obs =    203
                                                        Wald chi2(3)  =  36.80
                                                        Prob > chi2   = 0.0000
Log pseudolikelihood = -24.932672                       Pseudo R2     = 0.4530

                                     (Std. err. adjusted for 31 clusters in intid)
----------------------------------------------------------------------------------
                 |               Robust
early_withdrawal | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
-----------------+----------------------------------------------------------------
           fcoup |   5.494239   1.255989     4.37   0.000     3.032546    7.955932
           scoup |   3.841594   1.280495     3.00   0.003     1.331869    6.351318
       lbestdead |   .3106436   .1162352     2.67   0.008     .0828268    .5384605
           _cons |  -4.534741   .5912347    -7.67   0.000     -5.69354   -3.375942
----------------------------------------------------------------------------------

. eststo m2

. 
. * Model 3: Liberal democracies excluded + controls
. logit early_withdrawal fcoup scoup lbestdead coal1_state coal3_alliance coalsize ///
>     direct2 domdispute combat ltdistance lldistance logcap troops ///
>     newll dur time time2 time3 ///
>     if type_of_conflict != 1 & 2 & polity2 < 7, cluster(intid) nolog

Logistic regression                                   Number of obs =      203
                                                      Wald chi2(18) = 18945.33
                                                      Prob > chi2   =   0.0000
Log pseudolikelihood = -9.8292933                     Pseudo R2     =   0.7843

                                     (Std. err. adjusted for 31 clusters in intid)
----------------------------------------------------------------------------------
                 |               Robust
early_withdrawal | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
-----------------+----------------------------------------------------------------
           fcoup |   11.66206   2.515587     4.64   0.000     6.731601    16.59252
           scoup |    7.96128   2.229185     3.57   0.000     3.592158     12.3304
       lbestdead |   1.291324   .3926576     3.29   0.001      .521729    2.060918
     coal1_state |  -3.084052   4.069679    -0.76   0.449    -11.06048    4.892372
  coal3_alliance |  -1.933482   3.206386    -0.60   0.547    -8.217882    4.350918
        coalsize |  -.2381303   .4078129    -0.58   0.559    -1.037429    .5611684
         direct2 |   2.929282   2.369383     1.24   0.216    -1.714624    7.573188
      domdispute |   6.198307   3.382771     1.83   0.067    -.4318021    12.82842
          combat |   3.844217    4.79164     0.80   0.422    -5.547225    13.23566
      ltdistance |  -.0649681   .1702725    -0.38   0.703    -.3986961    .2687599
      lldistance |   .2361592   .3956997     0.60   0.551     -.539398    1.011716
          logcap |  -.7512323   .6414326    -1.17   0.242    -2.008417    .5059526
          troops |  -.7593029    3.04873    -0.25   0.803    -6.734704    5.216098
           newll |   1.232756   4.946892     0.25   0.803    -8.462974    10.92849
             dur |  -3.526674   .9386424    -3.76   0.000     -5.36638   -1.686969
            time |   .6657591   3.598081     0.19   0.853    -6.386349    7.717867
           time2 |   .3426093   1.018556     0.34   0.737    -1.653724    2.338943
           time3 |  -.0325532   .0782729    -0.42   0.677    -.1859652    .1208589
           _cons |  -7.720197   11.71301    -0.66   0.510    -30.67728    15.23689
----------------------------------------------------------------------------------
Note: 19 failures and 0 successes completely determined.

. eststo m3

. 
. * Model 4: Failed coup only (LD excluded + controls)
. logit early_withdrawal fcoup coal1_state coal3_alliance coalsize ///
>     direct2 domdispute combat ltdistance lldistance logcap troops ///
>     newll dur time time2 time3 ///
>     if type_of_conflict != 1 & 2 & polity2 < 7, cluster(intid) nolog

Logistic regression                                     Number of obs =    203
                                                        Wald chi2(16) =  79.38
                                                        Prob > chi2   = 0.0000
Log pseudolikelihood = -22.073534                       Pseudo R2     = 0.5157

                                     (Std. err. adjusted for 31 clusters in intid)
----------------------------------------------------------------------------------
                 |               Robust
early_withdrawal | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
-----------------+----------------------------------------------------------------
           fcoup |   5.442082   1.813634     3.00   0.003     1.887425    8.996738
     coal1_state |  -11.87726   4.707397    -2.52   0.012    -21.10359   -2.650931
  coal3_alliance |  -10.53192   4.294086    -2.45   0.014    -18.94817   -2.115665
        coalsize |  -1.262585   .5298705    -2.38   0.017    -2.301112   -.2240578
         direct2 |  -1.120413   1.259762    -0.89   0.374    -3.589502    1.348676
      domdispute |   2.107559   1.143887     1.84   0.065    -.1344192    4.349537
          combat |   .0314308   1.932214     0.02   0.987    -3.755638      3.8185
      ltdistance |   .1471798   .0803096     1.83   0.067    -.0102242    .3045838
      lldistance |  -.1065752   .2206694    -0.48   0.629    -.5390793    .3259288
          logcap |    -.24154   .3731403    -0.65   0.517    -.9728815    .4898015
          troops |   .6905457   .7881708     0.88   0.381    -.8542407    2.235332
           newll |  -.3101423   .9576477    -0.32   0.746    -2.187097    1.566813
             dur |  -1.443521   .4108068    -3.51   0.000    -2.248687    -.638354
            time |   .8160491   1.292694     0.63   0.528    -1.717584    3.349682
           time2 |   -.060432   .2919593    -0.21   0.836    -.6326616    .5117977
           time3 |     .00139   .0176063     0.08   0.937    -.0331177    .0358976
           _cons |   10.35439   6.675655     1.55   0.121    -2.729654    23.43843
----------------------------------------------------------------------------------
Note: 13 failures and 0 successes completely determined.

. eststo m4

. 
. * Model 5: Successful coup only (LD excluded + controls)
. logit early_withdrawal scoup coal1_state coal3_alliance coalsize ///
>     direct2 domdispute combat ltdistance lldistance logcap troops ///
>     newll dur time time2 time3 ///
>     if type_of_conflict != 1 & 2 & polity2 < 7, cluster(intid) nolog

Logistic regression                                     Number of obs =    203
                                                        Wald chi2(16) =  64.69
                                                        Prob > chi2   = 0.0000
Log pseudolikelihood = -33.402345                       Pseudo R2     = 0.2671

                                     (Std. err. adjusted for 31 clusters in intid)
----------------------------------------------------------------------------------
                 |               Robust
early_withdrawal | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
-----------------+----------------------------------------------------------------
           scoup |   2.192485   .8469854     2.59   0.010     .5324238    3.852546
     coal1_state |  -5.874814   6.494644    -0.90   0.366    -18.60408    6.854455
  coal3_alliance |  -5.699333   5.133778    -1.11   0.267    -15.76135    4.362688
        coalsize |  -.6192437   .6926884    -0.89   0.371    -1.976888    .7384006
         direct2 |   .2736583   1.552717     0.18   0.860     -2.76961    3.316927
      domdispute |   1.778127   1.478664     1.20   0.229    -1.120001    4.676255
          combat |   .7742676     2.1621     0.36   0.720     -3.46337    5.011905
      ltdistance |   .0184954   .0826584     0.22   0.823    -.1435121    .1805028
      lldistance |  -.0522886   .2032357    -0.26   0.797    -.4506232     .346046
          logcap |  -.4353526   .3517378    -1.24   0.216    -1.124746    .2540407
          troops |   .6039799   .7825701     0.77   0.440    -.9298293    2.137789
           newll |   1.513923    1.00546     1.51   0.132    -.4567433    3.484589
             dur |  -.7630874    .436406    -1.75   0.080    -1.618427    .0922526
            time |  -1.582272   1.333114    -1.19   0.235    -4.195126    1.030583
           time2 |   .4657916   .3869305     1.20   0.229    -.2925782    1.224161
           time3 |  -.0324792   .0290844    -1.12   0.264    -.0894836    .0245252
           _cons |   2.719294   8.728085     0.31   0.755    -14.38744    19.82603
----------------------------------------------------------------------------------

. eststo m5

. 
. * Model 6: Battle deaths only (LD excluded + controls)
. logit early_withdrawal lbestdead coal1_state coal3_alliance coalsize ///
>     direct2 domdispute combat ltdistance lldistance logcap troops ///
>     newll dur time time2 time3 ///
>     if type_of_conflict != 1 & 2 & polity2 < 7, cluster(intid) nolog

Logistic regression                                     Number of obs =    203
                                                        Wald chi2(16) = 128.12
                                                        Prob > chi2   = 0.0000
Log pseudolikelihood = -23.42674                        Pseudo R2     = 0.4860

                                     (Std. err. adjusted for 31 clusters in intid)
----------------------------------------------------------------------------------
                 |               Robust
early_withdrawal | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
-----------------+----------------------------------------------------------------
       lbestdead |   .8159871   .2793294     2.92   0.003     .2685116    1.363463
     coal1_state |   -3.36662   4.714669    -0.71   0.475     -12.6072    5.873961
  coal3_alliance |  -4.106462   4.509686    -0.91   0.363    -12.94528    4.732359
        coalsize |  -.4719294   .6503068    -0.73   0.468    -1.746507    .8026484
         direct2 |   1.532121   1.370773     1.12   0.264    -1.154544    4.218786
      domdispute |   3.174863   1.821212     1.74   0.081    -.3946476    6.744374
          combat |   1.910794   2.668837     0.72   0.474    -3.320031    7.141618
      ltdistance |  -.1183978   .1817541    -0.65   0.515    -.4746292    .2378337
      lldistance |   .6199725   .5650971     1.10   0.273    -.4875975    1.727542
          logcap |  -1.204067   .6122816    -1.97   0.049    -2.404117   -.0040171
          troops |   .5701855   .8787296     0.65   0.516    -1.152093    2.292464
           newll |    3.60623   2.077964     1.74   0.083    -.4665041    7.678963
             dur |  -1.342556   .4200838    -3.20   0.001    -2.165905    -.519207
            time |  -2.592554   1.673513    -1.55   0.121    -5.872579    .6874712
           time2 |   .8010188    .477778     1.68   0.094    -.1354089    1.737447
           time3 |  -.0534876    .034332    -1.56   0.119    -.1207771    .0138018
           _cons |  -10.19483    5.06091    -2.01   0.044    -20.11403    -.275629
----------------------------------------------------------------------------------

. eststo m6

. 
. * Model 7: Liberal democracies not excluded + controls
. logit early_withdrawal fcoup scoup lbestdead coal1_state coal3_alliance coalsize ///
>     direct2 domdispute combat ltdistance lldistance logcap troops ///
>     newll election dem dur time time2 time3 ///
>     if type_of_conflict != 1 & 2, cluster(intid) nolog

Logistic regression                                     Number of obs =    395
                                                        Wald chi2(20) = 308.63
                                                        Prob > chi2   = 0.0000
Log pseudolikelihood = -28.750495                       Pseudo R2     = 0.5899

                                     (Std. err. adjusted for 43 clusters in intid)
----------------------------------------------------------------------------------
                 |               Robust
early_withdrawal | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
-----------------+----------------------------------------------------------------
           fcoup |   8.195057   2.749574     2.98   0.003     2.805992    13.58412
           scoup |   5.560005   1.145277     4.85   0.000     3.315303    7.804707
       lbestdead |   .6356158   .2021762     3.14   0.002     .2393578    1.031874
     coal1_state |  -5.931836   2.280663    -2.60   0.009    -10.40185   -1.461819
  coal3_alliance |  -5.521405   2.451662    -2.25   0.024    -10.32657   -.7162357
        coalsize |   .0619642   .2665489     0.23   0.816     -.460462    .5843904
         direct2 |   1.778582   1.766218     1.01   0.314    -1.683141    5.240305
      domdispute |   8.617745   3.040353     2.83   0.005     2.658762    14.57673
          combat |   1.586245   1.772442     0.89   0.371    -1.887677    5.060167
      ltdistance |  -.0603023   .0885118    -0.68   0.496    -.2337823    .1131776
      lldistance |  -.4133393    .106343    -3.89   0.000    -.6217678   -.2049108
          logcap |   .1236373   .3384645     0.37   0.715    -.5397409    .7870154
          troops |   .4538182   .9459314     0.48   0.631    -1.400173     2.30781
           newll |  -.1191181   1.865335    -0.06   0.949    -3.775107    3.536871
        election |  -.3336752   .9323185    -0.36   0.720    -2.160986    1.493636
             dem |     .39714   1.065306     0.37   0.709    -1.690821    2.485101
             dur |   -2.24929   .5144853    -4.37   0.000    -3.257662   -1.240917
            time |   1.339697   1.053473     1.27   0.203    -.7250724    3.404467
           time2 |  -.0127551   .2039107    -0.06   0.950    -.4124127    .3869025
           time3 |  -.0049451   .0123781    -0.40   0.690    -.0292058    .0193156
           _cons |   .6183258   5.049099     0.12   0.903    -9.277726    10.51438
----------------------------------------------------------------------------------

. eststo m7

. 
. * Export models to LaTeX
. esttab m1 m2 m3 m4 m5 m6 m7 using "7_models.tex", ///
>     replace se star(* 0.1 ** 0.05 *** 0.01) ///
>     stats(N r2_p, labels("N" "Pseudo R²")) ///
>     booktabs label compress ///
>     title("Table 2: Early Withdrawal from Military Coalitions, 1950–2001")
(file 7_models.tex not found)
(output written to 7_models.tex)

. 
. 
. ********************************************************************************
. * 4. FIGURES 1-2: MARGINAL EFFECTS CALCULATIONS
. ********************************************************************************
. 
. /*
> Note: The "margins" commands below are run to calculate the substantive effects. The "marginsplot" com
> mands are commented out, as the final figures for the manuscript were created in R for enhanced visual
> ization using the results generated here. Please run the accompanying R script(s) separately to regene
> rate the two figures
> */
. 
. * Re-run Model 3 for margins
. logit early_withdrawal fcoup scoup lbestdead coal1_state coal3_alliance coalsize ///
>     direct2 domdispute combat ltdistance lldistance logcap troops ///
>     newll dur time time2 time3 ///
>     if type_of_conflict != 1 & 2 & polity2 < 7, cluster(intid) nolog

Logistic regression                                   Number of obs =      203
                                                      Wald chi2(18) = 18945.33
                                                      Prob > chi2   =   0.0000
Log pseudolikelihood = -9.8292933                     Pseudo R2     =   0.7843

                                     (Std. err. adjusted for 31 clusters in intid)
----------------------------------------------------------------------------------
                 |               Robust
early_withdrawal | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
-----------------+----------------------------------------------------------------
           fcoup |   11.66206   2.515587     4.64   0.000     6.731601    16.59252
           scoup |    7.96128   2.229185     3.57   0.000     3.592158     12.3304
       lbestdead |   1.291324   .3926576     3.29   0.001      .521729    2.060918
     coal1_state |  -3.084052   4.069679    -0.76   0.449    -11.06048    4.892372
  coal3_alliance |  -1.933482   3.206386    -0.60   0.547    -8.217882    4.350918
        coalsize |  -.2381303   .4078129    -0.58   0.559    -1.037429    .5611684
         direct2 |   2.929282   2.369383     1.24   0.216    -1.714624    7.573188
      domdispute |   6.198307   3.382771     1.83   0.067    -.4318021    12.82842
          combat |   3.844217    4.79164     0.80   0.422    -5.547225    13.23566
      ltdistance |  -.0649681   .1702725    -0.38   0.703    -.3986961    .2687599
      lldistance |   .2361592   .3956997     0.60   0.551     -.539398    1.011716
          logcap |  -.7512323   .6414326    -1.17   0.242    -2.008417    .5059526
          troops |  -.7593029    3.04873    -0.25   0.803    -6.734704    5.216098
           newll |   1.232756   4.946892     0.25   0.803    -8.462974    10.92849
             dur |  -3.526674   .9386424    -3.76   0.000     -5.36638   -1.686969
            time |   .6657591   3.598081     0.19   0.853    -6.386349    7.717867
           time2 |   .3426093   1.018556     0.34   0.737    -1.653724    2.338943
           time3 |  -.0325532   .0782729    -0.42   0.677    -.1859652    .1208589
           _cons |  -7.720197   11.71301    -0.66   0.510    -30.67728    15.23689
----------------------------------------------------------------------------------
Note: 19 failures and 0 successes completely determined.

. 
. * Marginal effects: Failed coup
. margins, at(fcoup = (0 1))

Predictive margins                                         Number of obs = 203
Model VCE: Robust

Expression: Pr(early_withdrawal), predict()
1._at: fcoup = 0
2._at: fcoup = 1

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .0301095   .0087095     3.46   0.001     .0130392    .0471799
          2  |    .562134   .0628211     8.95   0.000     .4390069     .685261
------------------------------------------------------------------------------

.  marginsplot

Variables that uniquely identify margins: fcoup

. 
. * Marginal effects: Successful coup
. margins, at(scoup = (0 1))

Predictive margins                                         Number of obs = 203
Model VCE: Robust

Expression: Pr(early_withdrawal), predict()
1._at: scoup = 0
2._at: scoup = 1

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .0562288   .0111512     5.04   0.000     .0343728    .0780847
          2  |   .3342671   .1086526     3.08   0.002     .1213119    .5472222
------------------------------------------------------------------------------

.  marginsplot

Variables that uniquely identify margins: scoup

. 
. * Marginal effects: Battle deaths
. margins, at(lbestdead = (0(1)13.11))

Predictive margins                                         Number of obs = 203
Model VCE: Robust

Expression: Pr(early_withdrawal), predict()
1._at:  lbestdead =  0
2._at:  lbestdead =  1
3._at:  lbestdead =  2
4._at:  lbestdead =  3
5._at:  lbestdead =  4
6._at:  lbestdead =  5
7._at:  lbestdead =  6
8._at:  lbestdead =  7
9._at:  lbestdead =  8
10._at: lbestdead =  9
11._at: lbestdead = 10
12._at: lbestdead = 11
13._at: lbestdead = 12
14._at: lbestdead = 13

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .0389623   .0094188     4.14   0.000     .0205017    .0574228
          2  |    .049677    .012196     4.07   0.000     .0257733    .0735806
          3  |   .0670591   .0219812     3.05   0.002     .0239767    .1101415
          4  |   .0942255   .0360916     2.61   0.009     .0234873    .1649638
          5  |   .1316905   .0508763     2.59   0.010     .0319748    .2314061
          6  |   .1809502   .0718769     2.52   0.012      .040074    .3218264
          7  |   .2444327   .0904018     2.70   0.007     .0672483     .421617
          8  |   .3184713   .0934261     3.41   0.001     .1353596    .5015831
          9  |   .3970322   .0821557     4.83   0.000     .2360099    .5580544
         10  |   .4756801   .0680483     6.99   0.000     .3423079    .6090523
         11  |   .5524003   .0687806     8.03   0.000     .4175927    .6872079
         12  |   .6283755   .0906283     6.93   0.000     .4507473    .8060038
         13  |   .7068548   .1123579     6.29   0.000     .4866373    .9270724
         14  |   .7873711   .1243624     6.33   0.000     .5436253    1.031117
------------------------------------------------------------------------------

.  marginsplot

Variables that uniquely identify margins: lbestdead

. 
. 
. ********************************************************************************
. * 5. APPENDIX: ROBUSTNESS CHECKS
. ********************************************************************************
. 
. eststo clear

. 
. * (A) Alternative DV windows – Non-consolidated sample
. logit alt_dv_1 fcoup scoup lbestdead coal1_state coal3_alliance coalsize ///
>     direct2 domdispute combat ltdistance lldistance logcap troops ///
>     newll dur time time2 time3 ///
>     if type_of_conflict != 1 & 2 & polity2 < 7, cluster(intid) nolog

Logistic regression                                 Number of obs =        203
                                                    Wald chi2(18) = 3052583.12
                                                    Prob > chi2   =     0.0000
Log pseudolikelihood = -7.4838967                   Pseudo R2     =     0.8122

                                   (Std. err. adjusted for 31 clusters in intid)
--------------------------------------------------------------------------------
               |               Robust
      alt_dv_1 | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
---------------+----------------------------------------------------------------
         fcoup |   10.30749   1.333639     7.73   0.000     7.693603    12.92137
         scoup |   17.90844   7.485752     2.39   0.017     3.236636    32.58024
     lbestdead |   1.489095   .2770663     5.37   0.000     .9460549    2.032135
   coal1_state |  -5.935026   20.65234    -0.29   0.774    -46.41287    34.54281
coal3_alliance |  -6.408484    20.9491    -0.31   0.760    -47.46797      34.651
      coalsize |   1.284734   .9694944     1.33   0.185    -.6154405    3.184908
       direct2 |   15.20354    3.80033     4.00   0.000     7.755029    22.65205
    domdispute |   28.95188   17.91306     1.62   0.106    -6.157071    64.06083
        combat |   4.622543   3.514122     1.32   0.188     -2.26501     11.5101
    ltdistance |   .0629051   .1534888     0.41   0.682    -.2379274    .3637376
    lldistance |  -3.013563   2.217263    -1.36   0.174    -7.359318    1.332192
        logcap |  -.8958719   .5110145    -1.75   0.080    -1.897442    .1056981
        troops |  -.9199488   3.090519    -0.30   0.766    -6.977254    5.137357
         newll |   8.799569   2.746223     3.20   0.001      3.41707    14.18207
           dur |  -3.555074   .6222403    -5.71   0.000    -4.774642   -2.335505
          time |  -38.80009   9.430594    -4.11   0.000    -57.28371   -20.31646
         time2 |   21.65025   4.254337     5.09   0.000      13.3119    29.98859
         time3 |  -3.481548   .6641063    -5.24   0.000    -4.783172   -2.179923
         _cons |   11.45166   23.60493     0.49   0.628    -34.81315    57.71646
--------------------------------------------------------------------------------
Note: 107 failures and 0 successes completely determined.

. eststo R1_dv2

. 
. logit alt_dv_2 fcoup scoup lbestdead coal1_state coal3_alliance coalsize ///
>     direct2 domdispute combat ltdistance lldistance logcap troops ///
>     newll dur time time2 time3 ///
>     if type_of_conflict != 1 & 2 & polity2 < 7, cluster(intid) nolog

Logistic regression                                     Number of obs =    203
                                                        Wald chi2(18) = 631.77
                                                        Prob > chi2   = 0.0000
Log pseudolikelihood = -16.619845                       Pseudo R2     = 0.6559

                                   (Std. err. adjusted for 31 clusters in intid)
--------------------------------------------------------------------------------
               |               Robust
      alt_dv_2 | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
---------------+----------------------------------------------------------------
         fcoup |   7.286183   1.294489     5.63   0.000      4.74903    9.823335
         scoup |   4.479229   1.847644     2.42   0.015      .857912    8.100545
     lbestdead |   .8746625   .3457046     2.53   0.011     .1970939    1.552231
   coal1_state |  -2.752727   2.573514    -1.07   0.285    -7.796721    2.291268
coal3_alliance |  -2.801915   2.247649    -1.25   0.213    -7.207227    1.603396
      coalsize |  -.6102944   .3463817    -1.76   0.078     -1.28919    .0686012
       direct2 |   1.010874   1.610663     0.63   0.530    -2.145968    4.167716
    domdispute |   1.145691   2.458287     0.47   0.641    -3.672463    5.963846
        combat |   1.301952    1.88697     0.69   0.490    -2.396442    5.000345
    ltdistance |  -.1761033   .2270132    -0.78   0.438     -.621041    .2688343
    lldistance |   .8544274   .6917628     1.24   0.217    -.5014028    2.210258
        logcap |  -.9566107   .4637694    -2.06   0.039    -1.865582   -.0476394
        troops |  -.0666943   1.110843    -0.06   0.952    -2.243907    2.110519
         newll |   1.013485   2.377544     0.43   0.670    -3.646415    5.673386
           dur |  -1.654641    .554682    -2.98   0.003    -2.741797    -.567484
          time |   .1353029   3.111303     0.04   0.965     -5.96274    6.233346
         time2 |   .3886874   .9846878     0.39   0.693    -1.541265     2.31864
         time3 |   -.049242   .0919716    -0.54   0.592    -.2295031    .1310191
         _cons |  -9.896742   6.977162    -1.42   0.156    -23.57173    3.778244
--------------------------------------------------------------------------------
Note: 3 failures and 0 successes completely determined.

. eststo R2_dv3

. 
. * (B) Alternative DV windows – Full sample (+ election/dem)
. logit alt_dv_1 fcoup scoup lbestdead coal1_state coal3_alliance coalsize ///
>     direct2 domdispute combat ltdistance lldistance logcap troops ///
>     newll election dem dur time time2 time3 ///
>     if type_of_conflict != 1 & 2, cluster(intid) nolog

Logistic regression                                     Number of obs =    395
                                                        Wald chi2(17) =      .
                                                        Prob > chi2   =      .
Log pseudolikelihood = -16.829131                       Pseudo R2     = 0.7219

                                   (Std. err. adjusted for 43 clusters in intid)
--------------------------------------------------------------------------------
               |               Robust
      alt_dv_1 | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
---------------+----------------------------------------------------------------
         fcoup |   6.872784    3.17355     2.17   0.030     .6527403    13.09283
         scoup |    6.86473    2.98071     2.30   0.021     1.022646    12.70681
     lbestdead |    1.11946   .5467435     2.05   0.041     .0478623    2.191057
   coal1_state |  -3.498304    2.33971    -1.50   0.135    -8.084051    1.087443
coal3_alliance |  -4.036834   2.397899    -1.68   0.092     -8.73663    .6629627
      coalsize |   .6659733   .4617416     1.44   0.149    -.2390236     1.57097
       direct2 |   4.503955   3.043319     1.48   0.139     -1.46084    10.46875
    domdispute |   13.72194   7.744296     1.77   0.076    -1.456602    28.90048
        combat |   4.488235   2.406544     1.87   0.062    -.2285052    9.204975
    ltdistance |  -.1211381   .1975289    -0.61   0.540    -.5082877    .2660115
    lldistance |  -.4848122   .1521937    -3.19   0.001    -.7831063    -.186518
        logcap |  -.0881252   .2634348    -0.33   0.738    -.6044479    .4281974
        troops |   1.033866   1.745797     0.59   0.554    -2.387833    4.455566
         newll |  -.9976229   1.888807    -0.53   0.597    -4.699616     2.70437
      election |   1.612286   1.228689     1.31   0.189    -.7959012    4.020473
           dem |    1.66631   2.385929     0.70   0.485    -3.010025    6.342646
           dur |  -2.763272   .8755258    -3.16   0.002    -4.479271   -1.047273
          time |  -41.81709   19.34592    -2.16   0.031     -79.7344   -3.899782
         time2 |   21.96724   8.277622     2.65   0.008       5.7434    38.19108
         time3 |  -3.439996   1.058634    -3.25   0.001     -5.51488   -1.365112
         _cons |   13.32014   10.53289     1.26   0.206    -7.323957    33.96423
--------------------------------------------------------------------------------
Note: 115 failures and 0 successes completely determined.

. eststo R3_dv2_full

. 
. logit alt_dv_2 fcoup scoup lbestdead coal1_state coal3_alliance coalsize ///
>     direct2 domdispute combat ltdistance lldistance logcap troops ///
>     newll election dem dur time time2 time3 ///
>     if type_of_conflict != 1 & 2, cluster(intid) nolog

Logistic regression                                     Number of obs =    395
                                                        Wald chi2(20) = 925.92
                                                        Prob > chi2   = 0.0000
Log pseudolikelihood = -39.32636                        Pseudo R2     = 0.4626

                                   (Std. err. adjusted for 43 clusters in intid)
--------------------------------------------------------------------------------
               |               Robust
      alt_dv_2 | Coefficient  std. err.      z    P>|z|     [95% conf. interval]
---------------+----------------------------------------------------------------
         fcoup |   5.901845   1.230999     4.79   0.000     3.489132    8.314559
         scoup |   4.141871    1.23875     3.34   0.001     1.713966    6.569777
     lbestdead |   .3675418   .2534161     1.45   0.147    -.1291446    .8642282
   coal1_state |  -3.287651   2.062803    -1.59   0.111     -7.33067     .755369
coal3_alliance |  -3.230765    2.00834    -1.61   0.108    -7.167039    .7055094
      coalsize |  -.0346217    .242538    -0.14   0.886    -.5099873     .440744
       direct2 |    1.22009   1.232983     0.99   0.322    -1.196512    3.636692
    domdispute |   4.708897   3.388243     1.39   0.165    -1.931936    11.34973
        combat |   .9313935   1.556425     0.60   0.550    -2.119144    3.981931
    ltdistance |   .0090382   .0685233     0.13   0.895    -.1252651    .1433415
    lldistance |  -.2818257   .1483263    -1.90   0.057    -.5725398    .0088885
        logcap |   .0635023   .2915417     0.22   0.828     -.507909    .6349135
        troops |   .1516345   .6370092     0.24   0.812    -1.096881     1.40015
         newll |  -.0487742    1.30646    -0.04   0.970    -2.609388     2.51184
      election |  -.3448025   .5386133    -0.64   0.522    -1.400465    .7108602
           dem |   -.288124   .8799831    -0.33   0.743    -2.012859    1.436611
           dur |  -1.425858   .6038421    -2.36   0.018    -2.609367   -.2423496
          time |   1.146048   1.059591     1.08   0.279    -.9307112    3.222808
         time2 |  -.0626519    .207053    -0.30   0.762    -.4684683    .3431646
         time3 |  -.0010452   .0112763    -0.09   0.926    -.0231464    .0210561
         _cons |   .1642395   3.022954     0.05   0.957    -5.760642    6.089121
--------------------------------------------------------------------------------

. eststo R4_dv3_full

. 
. * Export robustness results to LaTeX
. esttab R1_dv2 R2_dv3 R3_dv2_full R4_dv3_full using "robustness_models.tex", ///
>     replace se star(* 0.10 ** 0.05 *** 0.01) ///
>     stats(N r2_p, labels("N" "Pseudo R²")) ///
>     booktabs label compress ///
>     title("Appendix Table X: Robustness Checks (DV Windows, SEs, Full Sample)")
(file robustness_models.tex not found)
(output written to robustness_models.tex)

. 
. 
. ********************************************************************************
. * 6. END OF DO-FILE
. ********************************************************************************
. 
. * Close the log file
. log close
      name:  <unnamed>
       log:  /Users/weiss.mehrabi/Desktop/Replication/replication_log.log
  log type:  text
 closed on:   9 Nov 2025, 21:23:30
--------------------------------------------------------------------------------------------------------
